Building a faster search algorithm with intuitive machine learning using BM25 and Js

Search is key when building any application this days. You are likely to gather data and most times you need to throw that data back at users but the problem is that users are the ones who request for these data. And most times they search using natural language and not keywords. Having worked at ConnectNigeria as CTO in the past, I was able to build a search algorithm that delivered over 80% accuracy for all search queries. Even queries such as “where i fit chop confab amala and weed returned the right results”. I was also able to pull off an accurate “did you mean feature” which understood yoruba, igbo, hausa and foreign languages as well. It was fun working Ahmad Mukoshy at that time. We built in php(wonder…


Link to Full Article: Building a faster search algorithm with intuitive machine learning using BM25 and Js